Fig 1.
Cutin biosynthesis genes are potential targets of RGL2 and ABI5.
(A) For whole genome expression studies (microarray), WT (Col) and rgl2 seeds were harvested 10, 18, 24 and 36 hours after imbibition under low GA conditions (PAC). Dynamics of cuticle formation gene expression between rgl2 and WT seeds is represented with a color code. The scale bar relates color with absolute fold changes. Genes in bold are involved in cutin synthesis. (B) For the ChIP-Seq experiment, seeds of abi5-4/ ind:: HA-ABI5 were imbibed in the presence of estradiol and ABA and harvested 36 hours after imbibition [39]. The table shows cutin biosynthesis genes with high HA-ABI5 gene occupancy.
Fig 2.
A cuticle is present on the outer side of endosperm cells.
(A) Sudan Red staining of a Col-0 seed section. Bar: 100μm. (B) Close-up of a Col-0 seed section stained with Sudan Red. The pink line between the endosperm cell bodies and the testa (arrowheads) indicates the presence of a cuticle on the outer side of endosperm cells. This pink line is specific to the Sudan Red staining, as it is not seen in unstained seed sections (right panel). T: testa; En: endosperm; E: Embryo. Bars: 15μm. (C) Electron micrograph of a Col-0 endosperm cell. Bar: 1000nm. (D-G) Close-ups of the outer (D, F) and inner (E, G) sides of endosperm cells in Col-0 seeds (D, E) and bdg seeds (F, G). Note the presence of a cuticle exclusively on the outer side of endosperm cells. The arrows in (F) indicate defects in the cuticle of a bdg1 cell. Abbreviations are: T: testa; C: cuticle; PW: primary wall. Bars: 300nm. (H-I) Profiles of the cuticle optical intensity along the dotted lines in (D) and (F), respectively. (J-K) Profiles of cuticle optical intensity in all the regions analysed in Col-0 cells (J; 969 lines) and bdg1 cells (K; 1002 lines). See the Materials and Methodssections for details on the measurement procedure. (L) Average cuticle optical density in Col-0 cells (green line) and bdg1 cells (red line). Standard errors of the mean are represented as shaded areas around each line. The blue line represents (on a semi-log scale) the p-values given by Kolmogorov-Smirnov tests performed at each position along the cuticle between the Col-0 and the bdg1 cuticle densities; the dark blue dotted line represents the 10−2 threshold of significance for these p-values.
Fig 3.
Imbibed seeds of cutin biosynthesis mutants can rupture testa under low GA conditions.
(A) Images showing WT (Col), lacs2, lcr, bdg1, dcr, gpat4/8, gpat5 seeds 72h after imbibition under low GA conditions. Black arrows indicate testa rupture. (B) Chart represents percentages of testa rupture over time (in hours) under low GA conditions (4 replicates (n = 150–200)).
Fig 4.
Expression of cutin biosynthesis genes is low when GA synthesis is inhibited.
(A) Histograms show the relative BDG1, GPAT4 and LACS2 mRNA accumulation in WT (Col) dry seeds as well as 8h, 12h, 16h, 22h and 26 h (MS) and 10h, 20h, 24h, 30h and 36 h (PAC) and after imbibition under normal and low GA conditions (PAC).
Fig 5.
Testa rupture correlates with the expansion of the micropylar endosperm cells. bdg1 and lacs2 mutants fail to arrest micropylar endosperm cell expansion when GA synthesis is inhibited.
(A) Histograms show the average area of micropylar endosperm cells in WT (Col), lacs2, bdg1 seeds at 1 and 24 hours after imbibition under normal conditions (MS) and 1 and 120 hours after imbibition under low GA conditions (PAC). (*** = P-values < 0.001, n equal 80 to 220 cells). Bar: 200μm. (B) Histograms show the average area of peripheral endosperm cells in WT (Col), lacs2 and bdg1 seeds at 1and 24 hours after imbibition under normal conditions (MS) and 1and 120 hours after imbibition under low GA conditions (PAC). (*** = P-values < 0.001, n equal 70 to 150 cells). The same seeds used in (A) were used for peripheral endosperm measurements. Bar: 200μm. (C) Representative pictures of micropylar and peripheral endosperm cells of WT (Col), lacs2 and bdg1 taken at 1 and 24 hours after imbibition under normal conditions (MS). Yellow arrow indicates a cell of the endosperm. Bars: 10μm. (D) Same as (C) with pictures taken 1 and 120 hours after imbibition under low GA conditions (PAC).
Fig 6.
bdg1 endosperm exhibits higher permeability to toluidine blue.
(A) Comparison of toluidine blue staining in embryos dissected from WT and bdg1 seeds. Whole seeds of WT and bdg1 were incubated on MS, ABA and PAC for different times prior or after TR as indicated. Thereafter, seed material was transferred to a toluidine blue solution for 6 hours. The solution maintained PAC or ABA if these compounds were previously present in the germination medium. For illustration, a picture of a seed with or without TR is shown. Pictures show embryos dissected out of seed coats after 6h incubation in toluidine blue solution. Histograms show blue intensity quantification obtained from pictures (Materials and Methods). (*** = P-values < 0.001).
Fig 7.
Dry seeds of cutin biosynthesis mutants lose viability faster than WT.
An accelerated aging treatment is applied for 0, 2, 4 and 6 days on WT (Col), lacs2, bdg1, dcr, gpat4/8, gpat5 dry seeds. Graphs represent a percentage of seeds germination over time (in days) (seeds were plated in triplicate (n = 100–150)). Before imbibition dry seeds underwent an accelerated aging treatment for 0, 2 and 4 days (indicated on the graphs). Seeds were germinating at 15°C in darkness.
Fig 8.
Cutin biosynthesis mutants exhibit low dormancy levels.
(A) Germination is induced after a FR pulse (5min) is immediately followed by a R pulse (5min)(FR/R). WT (Col), lacs2, bdg1, gpat4/8 seeds with different after ripening times were germinating in darkness and pictures were taken 3 days after FR/R treatment. (B) Histograms represent the percentage of germination of WT (Col), lacs2, bdg1, lcr, dcr, gpat4/8, gpat5 seeds 3 days after imbibition under suboptimal germination conditions (seeds were plated in triplicate (n = 100–150)).
Fig 9.
Polyunsaturated fatty acid hydroxides and tocochromanols in Arabidopsis cutin-deficient seeds.
Seed lipid oxidation was measured by quantifying 9- and 13-hydroxy derivatives of linoleic and linolenic acids in untreated seeds (A) and seeds aged for 6 days (B). Data are expressed as pmol of hydroxy fatty acid per nmol of unmodified fatty acid (average±STDEV; n = 2). Tocochromanol levels in untreated seeds (average±STDEV; n = 2) (C). Asterisks represent significance levels using Student’s t test of each genotype relative to Col-0 WT controls. *: P<0.05; **: P<0.01).
Fig 10.
Expression of cutin biosynthetic genes is repressed by ABA under low GA conditions.
(A) Histograms show the relative BDG1, GPAT4 and LACS2 mRNA accumulation in dry seeds as well as upon seed imbibition under low GA conditions in WT (Col), rgl2 and aba1. (B) Same as (A) but with abi5 and snrk2.2/snrk2.3 mutants.
Fig 11.
Hypothetical role of a cutin-containing endospermic cuticle in seed physiology.
(A) A cuticle limits the diffusion of oxygen within seeds. In cutin biosynthesis mutants, the structure is deficient, which leads to higher diffusion of oxygen and thus higher accumulation of LOH. In turn, higher oxidative stress accelerates loss of dormancy and seed aging. (B) When GA synthesis is blocked, micropylar endosperm cell expansion and testa rupture are blocked. Under these conditions, de novo cutin biosynthesis gene expression is blocked. The cuticular structure already present in dry seeds is therefore maintained and participates to prevent cell expansion and testa rupture. Cell expansion involves water uptake within cells. The role assigned to the endospermic cuticle in this context is suggested by the fact that cutin biosynthesis mutants fail to repress cell expansion and testa rupture. Cl: cuticle, PW: primary cell wall.